Title: Learning-Based Image Rendering

Abstract: Image rendering is essential study field for computer science, robotics, and augmented reality. In the last decade, the increase in the graphics processing power of computers and the widespread use of deep learning networks have led to deep learning networks being at the heart of the studies on image rendering. The use of deeper networks improves the visual representation ability of the trained models and gives them the ability to render high quality images. In this study, various information is given about subjects such as image rendering, obtaining 3D data from 2D data, 3D image rendering, differentiable rendering and recent studies on this subject.

Keywords: computer vision; depth image; point cloud generation; differentiable rendering

Başlık: Öğrenme Tabanlı Görüntü İşleme

Özet: Görüntü işleme, bilgisayar bilimleri,robotik ve artırılmış gerçeklik gibi çalışma alanları için oldukça önemlidir. Son on yılda, işlem yeteneği ve işlem gücü artan GPU’lar sayesinde derin öğrenme ,yapay sinir ağları ve görüntü işleme konuları aynı çalışmaların içinde daha fazla kullanılmaya başlanmıştır.Artan işlem gücü ile birlikte daha derin ve kompleks modeller eğitilebilmektedir.Bu çalışmada görüntü işleme,2B veriden 3B veri elde etmek , 3B görüntü işleme ve türevlenebilir işleme gibi konular hakkında ve son zamanlarda bu konu ile ilgili yapılan çalışmalar hakkında çeşitli bilgiler verilmektedir.

Anahtar kelimeler: bilgisayar grafikleri; derinlik; nokta bulutu üretme; türevlenebilir işleme

• Kutlu Y, Alanoglu Z, Gokcen A, Yeniad M. Raspberry Pi based intelligent robot that recognizes and places puzzle objects. Journal of Intelligent Systems with Applications 2019; 2(1): 85-89.
• Yasar A, Uslu E, Cakmak F, Altuntas N, Amasyali MF, Yavuz S. Autonomous mobile robot navigation in structured rough terrain. Journal of Intelligent Systems with Applications 2018; 1(1): 67-74
• Uslu T, Cetin L, Gezgin E. Preliminary study of a surgical navigation with point based registration method. Journal of Intelligent Systems with Applications 2021; 4(1): 6-9.
• Kato H, Beker D, Morariu M, Ando T, Matsuoka T, Kehl W, Gaidon A. Differentiable rendering: A survey. arXiv preprint 2020; arXiv:2006.12057.
• Zhang J, Ramanagopal MS, VasudevanR, Johnson-Roberson M. Listereo: Generate dense depth maps from lidar and stereo imagery. In IEEE International Conference on Robotics and Automation (ICRA), May 31-August 31, 2020, Paris, France , pp. 7829-7836.
• Imran S, Long Y, Liu X, Morris D. Depth coefficients for depth completion. In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 15-20, 2019, Long Beach, CA, USA, pp. 12438-12447.
• Zhang Y, Funkhouser T. Deep depth completion of a single RGB-D image. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 18-23, 2018, Salt Lake City, UT, USA, pp. 175-185.
• Shinohara T, Xiu H, Matsuoka M. Point2color: 3D point cloud colorization using a conditional generative network and differentiable rendering for airborne LiDAR. In IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), June 19-25, 2021, Nashville, TN, USA, pp. 1062-1071.
• Zhou QY, Park J, Koltun V. Open3D: A modern library for 3D data processing. arXiv preprint 2018; arXiv:1801.09847.
• Luo S, Hu W. (2021). Diffusion probabilistic models for 3D point cloud generation. In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 20-25, 2021, Nashville, TN, USA, pp. 2837-2845,
• Mildenhall B, Srinivasan PP, Tancik M, Barron JT, Ramamoorthi R, Ng R. Nerf: Representing scenes as neural radiance fields for view synthesis. In 16th European conference on computer vision, August 23-28, 2020, pp. 405-421.
• Wiles O, Gkioxari G, Szeliski R, Johnson J. Synsin: End-to-end view synthesis from a single image. In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 13-19, 2020, Seattle, WA, USA, pp. 7467-7477.
• Ravi N, Reizenstein J, Novotny D, Gordon T, Lo WY, Johnson J, Gkioxari G. Accelerating 3D deep learning with pyTorch3D. arXiv preprint 2020; arXiv:2007.08501.